This research is aimed at developing a reliability model using values of predicted cohesion with increase in liquid limits, plastic limits, effective unit weight and depth of soil geotechnical parameters from laboratory field investigation at Assa Imo state. The model involves the use of Least Square method in simulating and analyzing data collated from the study area using Python and Microsoft excel tool. Cohesion (Cu) and Angle of internal friction (φ) were a function of depth (d), liquid limit (LL), plastic limit (PL) and unit weight (γ) from field test results, which were used to stimulate and predict new Cu and φ values at arbitrary depth in the model. The model established a reliable prediction of soil parameters for design purpose using existing field data of vertical spatial variable soil parameters form the study area, as the study indicates a direct relationship between increasing values of liquid limit (LL), plastic limit (PL), effective unit weight (γ), and depth, and the observed enhancements in cohesion (Cu) and the angle of internal friction (φ). The model demonstrates that changes in the liquid limit, plastic limit, effective unit weight, and depth significantly affect the predicted values of Cu. At a depth of 5.5 metres, the recorded values are as follows: Liquid Limit (LL) is 55, Plastic Limit (PL) stands at 27.5, and the unit weight (γ) is 18.25 kN/m³, resulting in cohesion (Cu) of 85.8 KN/m2 and an Internal friction angle (φ) of 3.59 degrees. Upon increasing the depth to 6 metres, the measurements change to: LL at 60, PL at 30.0, and γ at 19.0 kN/m³, leading to a cohesion (Cu) of 97.2 KN/m2 and an internal friction angle (φ) of 4.61 degrees, this trend continued in the model as vertical increase in depth showed significant increase in Cu and φ values. The model was validated using three Goodness of fit tests, as all test shows “Good” and “Very Good” in the band of qualitative interpretation, as the correlation coefficient recorded 68.4% and 65.2% for predicted Cu and φ values respectively validating the reliability of the model. It is therefore possible to predict reliable soil parameters of depths not investigated for design purpose using available field data of known depths from a study area.
Tamunoemi Alu LongJohn (Wed,) studied this question.